Influence of factors related to electrode array placement on the early speech recognition of cochlear implant recipients with and without functional hearing preservation

For the hundreds of thousands of people worldwide who rely on cochlear implants to hear, outcomes have always been stubbornly variable. Two recipients implanted on the same day, with the same device and the same surgical team, can walk away with dramatically different speech recognition scores. Now a team at the University of North Carolina at Chapel Hill has traced part of that variability to something surprisingly physical: exactly where the electrode array ends up inside the cochlea, and how far it sits from the region of functional acoustic hearing that the brain still uses.

The study, published in Frontiers in Audiology and Otology, reviewed 34 adults who were candidates for hearing preservation cochlear implantation, meaning they still had usable low-frequency hearing in the ear to be implanted. All received a short, 23-millimeter straight electrode array, the Slim J version 2 from Advanced Bionics, chosen specifically to limit surgical trauma and maximize the odds that delicate residual hearing survives the operation. When that hearing is preserved, recipients are fitted with an electric-acoustic stimulation, or EAS, device that combines amplification of low frequencies with electrical stimulation of high frequencies. When it is lost, they fall back on a conventional cochlear implant alone.

The researchers zeroed in on two geometric factors. The first is frequency-to-place mismatch, the discrepancy between the frequency information that the device assigns to each electrode channel and the cochlear place frequency that the electrode actually stimulates. Because default mapping software does not account for an individual’s angular insertion depth, these mismatches are nearly universal. In this sample, the most apical electrode contact ranged from 279 to 512 degrees of insertion depth, a spread of more than 200 degrees across recipients, and the mismatch at the 1.5-kilohertz cochlear region ranged from 1 to 18 semitones. Twelve semitones is a full octave, meaning the brain receives pitch information an octave away from where the cochlea would naturally encode it.

The second factor is proximity, a measure specific to EAS users. It characterizes the angular insertion depth of the most apical electrode relative to the region of functional acoustic hearing, defined here as thresholds of 70 decibels hearing level or better, the crossover the reviewed EAS system uses to decide which frequencies are presented acoustically and which electrically. Three scenarios are possible: the array sits basal to the acoustic region, potentially leaving a stretch of unstimulated cochlea, a so-called stimulation gap; it sits right next to the region; or one or more electrodes land within it.

The results split cleanly by device type. Among the 15 patients who lost functional acoustic hearing and listened with a cochlear implant alone, word recognition in quiet at three months post-activation was significantly poorer as the magnitude of frequency-to-place mismatch grew, with a correlation of 0.61 and a p-value of 0.015. Sentence recognition in noise, however, showed no significant relationship with mismatch. The finding aligns with a growing body of evidence that mismatch matters early on, and it is clinically consequential because hearing-preservation candidates are deliberately given shorter arrays at shallower depths, which tends to increase mismatch.

For the 18 ears that retained aidable hearing and used an EAS device, the picture inverted. In 14 of them, the electrode array was placed entirely outside the functional acoustic hearing region; in only four did one or more contacts land within it. For the majority group, speech recognition in noise was not significantly associated with frequency-to-place mismatch, but it was dramatically associated with proximity, with a correlation of 0.96 and a p-value below 0.001. EAS users whose arrays sat farther from the acoustic hearing region scored markedly worse on the AzBio sentence test in a 10-talker babble at a 10-decibel signal-to-noise ratio. The pattern held during leave-one-out validation, with correlation coefficients ranging from 0.94 to 0.98.

The mechanism the authors favor is the stimulation gap. When the most apical electrode is far from the acoustic region, a stretch of the cochlea goes unstimulated, and current clinical mapping conventions do not account for it. EAS programming typically limits spectral gaps in the frequency output of the acoustic and electric components by dividing or overlapping frequency information, a practice based on evidence that spectral gaps hurt performance. But simulation studies suggest that gaps may not matter much when the implant’s filter frequencies match cochlear tonotopy, hinting that the physical gap between electrode and acoustic region may be the more important variable.

The discrepancy on mismatch for EAS users is itself intriguing. Previous work, including simulation studies with normal-hearing listeners and a longitudinal study of EAS users, found that larger mismatches slowed improvement over time. The present sample showed no such relationship, which the authors attribute cautiously to the small size and retrospective design, or to the possibility that proximity simply dominates early outcomes when the array sits outside the acoustic region. Only nine EAS users completed the in-noise assessment, and only four had electrodes within the acoustic region, too few to explore electric-on-acoustic masking, the interaction that occurs when electrical stimulation overlaps and potentially masks the acoustic cues the brain is trying to use.

The limitations are real. Uncorrected p-values were reported given the exploratory nature, raising the risk of Type 1 error across multiple comparisons. Patients with poor performance in quiet are typically not tested in noise, thinning the data. And the study captured only the first three months of device use, a period after which cochlear implant users generally plateau but EAS users may not. Still, the correlations are striking enough that the authors argue frequency-to-place mismatch and proximity deserve consideration in device selection and mapping, and investigations are already underway on mapping procedures that account for an individual’s electrode placement in filter frequency assignments and per-channel stimulation levels.

What emerges is a picture of cochlear implantation as a problem of geometry as much as biology. The same electrode array can land at wildly different depths depending on array design, surgical approach, and the twists of an individual’s cochlea, and the brain must reconcile whatever pitch information it receives with the tonotopic map it was born with. For the growing population of hearing-preservation candidates, who stand to benefit most from combining what remains of their natural hearing with electrical stimulation, getting the placement right, and then mapping to match it, may be the difference between struggling in a noisy restaurant and following the conversation with ease.

Subject of Research: Influence of factors related to electrode array placement on the early speech recognition of cochlear implant recipients with and without functional hearing preservation

Article Title: Influence of factors related to electrode array placement on the early speech recognition of cochlear implant recipients with and without functional hearing preservation

Article References: Dillon, M. T., Broome, M. E., Richter, M. E., Thompson, N. J., Lee, S. E., Scharf, S. P., Overton, A. B., Roth, N. E., Selleck, A. M., Dedmon, M. M., & Brown, K. D. (2026). Influence of factors related to electrode array placement on the early speech recognition of cochlear implant recipients with and without functional hearing preservation. Frontiers in Audiology and Otology, 4, Article 1908993. https://doi.org/10.3389/fauot.2026.1908993

Image Credits: AI Generated

DOI: 10.3389/fauot.2026.1908993

Keywords: Influence, factors, related, electrode, array, placement, early, speech, recognition, cochlear, implant, recipients